PROJECT SUMMARY The development of a preventive HIV-1 vaccine is a global health priority. If present before exposure to HIV-1, broadly neutralizing antibodies (bnAbs) can provide sterilizing immunity. Recent progress has been made in developing priming immunogens that successfully engage HIV-1 bnAb precursors. However, vaccine strategies to effectively guide maturation from these starting points to full neutralization breadth have been unsuccessful. Thus, there is an unmet need for new approaches to immunogen design. Drs. Bonsignori and Wiehe have pioneered the measurement of antibody mutation probabilities and showed that induction of bnAbs can be hindered by critical mutations. Critical mutations are defined as mutations that are either beneficial to develop neutralization breadth but occur with low probability, or detrimental to breadth and occur with high probability. Here, we propose a novel conceptual framework in which immunogens are designed to modulate the selection of critical mutations. We hypothesize that to elicit bnAbs, immunogens must select for beneficial improbable mutations and select against detrimental probable mutations. To test our hypothesis, we choose as a model the potent CD4-binding site VRC01-class of bnAbs. The VRC01 bnAb class is an appealing target for vaccine development due to its predictable immunogenetics and evidence of convergent evolution in multiple individuals. However, the challenges to VRC01-class bnAb induction are numerous. In particular, how to effectively accommodate the HIV-1 Env N276 glycan, an established hurdle to VRC01-class bnAb maturation, remains unresolved. The VRC01 lineage encompasses three different barriers to breadth that can be modulated by targeting critical mutations: early immuno-steering toward broadly neutralizing clades, insertions in the CDR H3 or deletions in the CDR L1. Hence, the VRC01 lineage is well-suited to validate our approach in these three scenarios. In Aim 1, we will identify the precise set of early beneficial mutations in broadly neutralizing clades and detrimental mutations in clades with limited neutralization. We will design immunogens to select for beneficial mutations and select against detrimental mutations using both a computational structural design approach and cell surface display. In Aims 2 and 3, we will deal with the N276 glycan barrier. Based on the observation that activation-induced cytidine deaminase (AID) hot spots regulate the frequency of insertions and deletions, in Aim 2, we will identify mutations that introduce AID hot spots in or around CDR L1 and design immunogens that will select for these mutations. We have demonstrated that VRC01 lineage bnAbs with long CDR H3s circumvented the N276 glycan barrier without mutating CDR L1. Thus, in Aim 3 we will design immunogens 1) to select for AID hot spots in or around CDR H3 and promote insertions, and 2) to select for VRC01-class bnAb precursors that have long CDR H3s. For each aim we will follow an in silico>in vitro>in vivo design strategy using a knock-in mouse model to test the ability of our designs to modulate critical mutations. If successful, this strategy can be broadly applied to other antibody specificities and pathogens.